Beyond security surveillance, video analytics and AI are converging to form the next frontier in data-driven operations across industries.
Across Asia Pacific, video technology is rapidly evolving beyond its traditional role in surveillance. With advances in AI, sensor integration and metadata analytics, video is emerging as a powerful operational data source that helps organizations understand real-world environments in real time.
This shift is driving the rise of “situational intelligence”, where video streams are transformed into actionable insights for improving safety, optimizing logistics, enhancing customer experience and managing complex infrastructure.
As adoption grows, open platforms, responsible AI and stronger data governance are becoming critical to unlocking the full value of video-driven operations.
Dr Barry Norton, Fellow at Milestone Systems, discusses how AI-powered video analytics is reshaping data-driven operations, the importance of openness in Asia’s diverse technology landscape, and what the next generation of intelligent video systems will look like.
Video technology has traditionally been viewed as a security tool. How is the role of video changing as organizations become more data-driven?
Dr Barry Norton: As organizations embrace digital transformation, video is expanding far beyond its historical role in surveillance. What’s changing is the entire ecosystem around surveillance devices. With AI, metadata extraction and multi-sensor fusion, video becomes a single, rich data channel capable of informing a wide array of operational decisions, helping us learn from the past, understand the present, and predict the future.
Beyond security, it is now just as relevant to operations, safety, planning and customer experience. In logistics, for instance, video analytics provide real-time visibility of disruptions and dwell times.
In manufacturing, video supports quality control and predictive maintenance. In retail, it informs customer flow optimization and the effectiveness of merchandising. In smart cities, it enhances traffic management, environmental monitoring and emergency response. In all these cases, video enables organizations to see patterns that were previously invisible.
The other major shift is towards automation. Instead of video being reviewed manually, analytics can automatically detect conditions, correlate them with other systems and trigger appropriate actions. As organizations become more data-driven, video becomes the most powerful data source, helping bridge the gap between digital and physical operations and supporting continuous improvement across multiple functions.
Why is an open-platform approach so important for unlocking the full value of video data and AI analytics in Asia’s diverse markets?
Dr Norton: Asia is one of the most diverse regions in the world — both culturally and technologically. Organizations here operate across a wide spectrum of infrastructures, regulatory requirements, levels of digital maturity and vendor ecosystems. In such an environment, no single vendor can realistically provide every capability needed for data-driven video operations. An open-platform approach is the only model that scales with this complexity.
Openness ensures customers can choose best-of-breed analytics, integrate with local or industry-specific solutions, and evolve their capabilities over time without vendor lock-in. It also allows the ecosystem to innovate faster. Analytics companies, device manufacturers and integrators can build on a common foundation rather than reinventing the stack for each project.
For Milestone, openness is a strategic enabler. It allows customers to harmonize cameras and other sensors, leverage emerging AI tools, comply with local governance constraints and adapt to future technologies without disrupting core operations.
In Asia, where requirements vary dramatically between a smart port, a manufacturing plant and a retail group, an open platform ensures video systems can be tailored, extended and future-proofed in a way that proprietary systems simply cannot deliver.
What major trends are you seeing in AI-powered video analytics, particularly around real-time detection, behavior analysis and predictive insights?
Dr Norton: AI-powered video analytics are rapidly evolving from detection of fixed objects and events to a more holistic understanding of behavior and context.
One major trend is multimodal fusion — combining video with physical and environmental sensors, location data and even textual or audio inputs. This enables more accurate, real-time decisioning and reduces the burden of false positives, especially in busy or complex environments.
Behavioral analytics is maturing. Instead of simply detecting object presence and movement, systems can now understand complex actions and interactions: unusual dwell patterns, deviations from normal flow, abnormal uses of equipment, or early indicators of safety risks. These models can rely on unsupervised or self-learning techniques, allowing them to adapt to the unique rhythms of each environment without extensive manual configuration.
Predictive analytics is the next frontier. By analyzing long-term metadata patterns, organizations can anticipate risks before they manifest — equipment failures, overcrowding, traffic congestion or safety hazards. The shift from reactive alerts to proactive guidance is a major step toward true operational intelligence.
We are now entering the era of Agentics, where AI systems can apply generative AI to understand context, make decisions, and take actions reflecting real user intent, rather than mechanically following prescribed steps. This marks a fundamental shift from ruled-based response from closed analytics components to proactive incident management.
By deploying AI agents, security leaders gain a technology teammate that can handle complexity autonomously, freeing professionals to focus on strategic judgement rather than operational reaction.
As video becomes a richer data source, concerns around privacy, governance and responsible use increase. What should be done to address responsible AI and data stewardship?
Dr Norton: Responsible handling of video data is fundamental to sustaining trust between organizations and regulators, and with the public at large.
Milestone’s approach begins with privacy-by-design, ensuring capabilities such as anonymization, access controls, encrypted communication and auditability are built into the platform from the outset rather than added as an afterthought.
On the AI side, we must be committed to transparency and explainability. Customers need to understand how systems arrive at a decision, what data they rely on and how models are validated. This includes clear documentation, audit trails and mechanisms to review or challenge automated decisions. An open-platform approach also plays a role here by allowing customers to choose AI tools that meet their governance requirements without being forced into proprietary models.
Data minimization is another core principle. Ensuring video and metadata are used for legitimate purposes, retained only as long as necessary and processed in accordance with regional regulations. Across Asia, where data governance standards vary widely, flexibility and clarity are key.
Ultimately, transparent technology and responsible AI are strategic differentiators. Organizations adopting AI must do so in a way that protects privacy, ensures fairness and builds long-term trust in how video data is used.
Many industries in Asia are investing in video for operational intelligence. What use cases are gaining the most traction?
Dr Norton: Operational intelligence is becoming a major driver of video investment across Asia, and we’re seeing strong demand in several sectors. In manufacturing, video analytics integrated with IoT sensors are used for quality assurance, safety monitoring and predictive maintenance, dramatically reducing downtime and enhancing compliance. Logistics facilities increasingly rely on video-driven situational awareness to optimize workflows, manage congestion and improve worker safety.
In retail, video supports heat-mapping, customer flow optimization and loss-prevention strategies that combine operational efficiency with improved customer experience. Smart cities continue to expand their reliance on video for real-time traffic optimization, environmental monitoring, public safety and infrastructure management.
Critical infrastructure is another high-growth area. Airports, ports and utilities use video to orchestrate large, complex operations where situational intelligence is essential for risk mitigation and continuity. Sectors such as healthcare and education are increasingly adopting video analytics for occupancy management, emergency response and service efficiency.
The common thread is that organizations are no longer buying cameras for security purposes alone. They are investing in a video intelligence layer that provides context; AI transforms it into insights; and integrated systems turn those insights into actionable decisions that improve performance, safety and resilience.
Looking ahead, what innovations or shifts will define the next generation of data-driven video solutions?
Dr Norton: The next generation of video solutions will be defined by three major shifts: autonomy, augmentation and convergence.
First, we will see more autonomous systems where AI agents orchestrate workflows end-to-end — aggregating data, generating insights and initiating operational actions with human oversight. This evolution will move organizations from reactive monitoring to continuous optimization.
Second, augmentation will play a larger role. Operators and field personnel will increasingly interact with video systems through natural-language interfaces, via text or voice, augmented-reality overlays and context-aware assistants. This will reduce cognitive load and bring insights directly to the point of action, not just the control room.
Third, the trend towards convergence will accelerate. Video will unify with building management, safety systems, robotics, IoT and enterprise platforms to create holistic digital twins. These present a persistent, real-time model of the environment that can be queried, simulated and automated.
At Milestone, we see a future where video becomes the connective link between people, processes and physical environments and we are building tools that will enable organizations to operate with greater clarity, agility and confidence.
